mirror of
https://github.com/microsoft/semantic-kernel.git
synced 2026-10-12 11:19:56 +00:00
Evan Mattson
8b16af69c0
### Motivation and Context As we move toward v1, it is necessary to remove the single kernel memory attribute so that memory connectors can be handled via plugins. This change also refactors the `import_plugin` method name to be similar with dotnet as `import_plugin_from_object`, and introduces a simple `kernel.invoke_prompt` method. <!-- Thank you for your contribution to the semantic-kernel repo! Please help reviewers and future users, providing the following information: 1. Why is this change required? 2. What problem does it solve? 3. What scenario does it contribute to? 4. If it fixes an open issue, please link to the issue here. --> ### Description Updates are: - The memory connector is now passed to the `SemanticTextMemory` class, along with the embedding generator. This allows the user to add memory to multiple plugins/services as needed. Closes #4632 - Kernel Examples and Notebooks were updated to reflect these changes. Now using Service Enum for all notebooks. Closes #4557 - There were planners using memory related to the kernel, to store the functions, and many of these planners are non-existent in dotnet's v1 SDK. Python has pending work to update the planner/remove deprecated planners before getting to v1. - A Kernel method called `invoke_prompt` was introduced to invoke a simple prompt, and return the `FunctionResult` when complete. - All unit/integration tests are passing. <!-- Describe your changes, the overall approach, the underlying design. These notes will help understanding how your code works. Thanks! --> ### Contribution Checklist <!-- Before submitting this PR, please make sure: --> - [X] The code builds clean without any errors or warnings - [X] The PR follows the [SK Contribution Guidelines](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md) and the [pre-submission formatting script](https://github.com/microsoft/semantic-kernel/blob/main/CONTRIBUTING.md#development-scripts) raises no violations - [X] All unit tests pass, and I have added new tests where possible - [ ] I didn't break anyone 😄
91 lines
3.2 KiB
Python
91 lines
3.2 KiB
Python
# Copyright (c) Microsoft. All rights reserved.
|
|
|
|
import os
|
|
|
|
from dotenv import load_dotenv
|
|
|
|
import semantic_kernel as sk
|
|
from semantic_kernel.connectors.ai.open_ai import OpenAIChatCompletion
|
|
from semantic_kernel.connectors.search_engine import GoogleConnector
|
|
from semantic_kernel.core_plugins import WebSearchEnginePlugin
|
|
|
|
load_dotenv()
|
|
|
|
|
|
async def main():
|
|
kernel = sk.Kernel()
|
|
api_key, org_id = sk.openai_settings_from_dot_env()
|
|
kernel.add_service(
|
|
OpenAIChatCompletion(service_id="chat-gpt", ai_model_id="gpt-3.5-turbo", api_key=api_key, org_id=org_id)
|
|
)
|
|
|
|
"""
|
|
Instantiate a Google Connector
|
|
Make sure to have the following keys in a .env file or set as environment variables
|
|
- GOOGLE_API_KEY
|
|
- GOOGLE_SEARCH_ENGINE_ID
|
|
|
|
A Google Custom Search API has to be created in order to have an API key and a search engine ID.
|
|
To create a Google Custom Search API, follow the guide - https://developers.google.com/custom-search/v1/overview.
|
|
If you have already created the service, the credentials can be found in the Credentials tab on the page
|
|
https://console.cloud.google.com/apis/api/customsearch.googleapis.com
|
|
"""
|
|
connector = GoogleConnector(
|
|
api_key=os.getenv("GOOGLE_API_KEY"),
|
|
search_engine_id=os.getenv("GOOGLE_SEARCH_ENGINE_ID"),
|
|
)
|
|
|
|
# Import the WebSearchEnginePlugin and pass the Google Connector to it.
|
|
web_plugin = kernel.import_plugin_from_object(WebSearchEnginePlugin(connector), "WebSearch")
|
|
|
|
# The search query
|
|
prompt = "Who is Leonardo DiCaprio's current girlfriend?"
|
|
search = web_plugin["searchAsync"]
|
|
|
|
# By default, only one search result is provided
|
|
result = await search.invoke(prompt)
|
|
print(result)
|
|
|
|
"""
|
|
Output:
|
|
["Celebrity Celebrity News Everything You Need to Know About Leonardo DiCaprio and Camila Morrone's
|
|
Relationship From the beginning of their romance to today, we track their relationship here. By..."]
|
|
"""
|
|
|
|
# Following example demonstrates the use of the plugin within a semantic function
|
|
prompt = """
|
|
Answer the question using only the data that is provided in the data section.
|
|
Do not use any prior knowledge to answer the question.
|
|
Data: {{WebSearch.SearchAsync "What is semantic kernel?"}}
|
|
Question: What is semantic kernel?
|
|
Answer:
|
|
"""
|
|
|
|
qna = kernel.create_semantic_function(prompt, temperature=0.2)
|
|
context = kernel.create_new_context()
|
|
|
|
"""
|
|
Two context parameters can be passed to the search engine plugin.
|
|
- num_results controls the number of results returned by the web search.
|
|
- offset controls the number of results to omit.
|
|
"""
|
|
context["num_results"] = "10"
|
|
context["offset"] = "0"
|
|
|
|
result = await qna.invoke(context=context)
|
|
print(result)
|
|
|
|
"""
|
|
Output:
|
|
Semantic Kernel is an open-source SDK that lets you easily combine AI services like OpenAI,
|
|
Azure OpenAI, and Hugging Face with conventional programming languages like C# and Python.
|
|
By doing so, you can create AI apps that combine the best of both worlds.
|
|
Semantic Kernel is at the center of the copilot stack.
|
|
"""
|
|
|
|
|
|
if __name__ == "__main__":
|
|
import asyncio
|
|
|
|
asyncio.run(main())
|